Polygenic prediction of body mass index and obesity through the life course and across ancestries

Abstract

Polygenic scores (PGSs) for body mass index (BMI) may guide early prevention and targeted treatment of obesity. Using genetic data from up to 5.1 million people (4.6% African ancestry, 14.4% American ancestry, 8.4% East Asian ancestry, 71.1% European ancestry and 1.5% South Asian ancestry) from the GIANT consortium and 23andMe, Inc., we developed ancestry-specific and multi-ancestry PGSs. The multi-ancestry score explained 17.6% of BMI variation among UK Biobank participants of European ancestry. For other populations, this ranged from 16% in East Asian-Americans to 2.2% in rural Ugandans. In the ALSPAC study, children with higher PGSs showed accelerated BMI gain from age 2.5 years to adolescence, with earlier adiposity rebound. Adding the PGS to predictors available at birth nearly doubled explained variance for BMI from age 5 onward (for example, from 11% to 21% at age 8). Up to age 5, adding the PGS to early-life BMI improved prediction of BMI at age 18 (for example, from 22% to 35% at age 5). Higher PGSs were associated with greater adult weight gain. In intensive lifestyle intervention trials, individuals with higher PGSs lost modestly more weight in the first year (0.55 kg per s.d.) but were more likely to regain it. Overall, these data show that PGSs have the potential to improve obesity prediction, particularly when implemented early in life.

Fig. 1. Study overview.

PGSs were constructed using ancestry-specific GWAS summary statistics, using ancestry-specific (PRS-CS) and ancestry-combined (PRS-CSx) approaches. Tuning of the global shrinkage parameter (ϕ) and optimal weights for the linear combination version of PRS-CSx was performed in the UKBB. The best-performing score across multiple ancestries was taken forward to independent validation studies (linear combination version of PRS-CSx with ϕ 1 × 10⁻²). Population descriptors shown in the figure reflect a combination of self-identified ethnicity and genetic similarity. Created in BioRender: Smit, R. (2025): https://BioRender.com/fglrflj.

Fig. 2. Explained variance for BMI in adult populations.

a, Explained variance for BMI in UKBB tuning populations, defined as adjusted R² of the rank-based inverse-normal transformed BMI (by sex) predicted by the PGS, incremental to age, genotyping array and ancestry principal components. Error bars represent 95% confidence intervals (CIs) from 1,000 bootstrap resamples. The ancestry-matched PGS is the best-performing PRS-CS score using ancestry-specific GWAS summary statistics. The genome-wide significant score reflects a weighted sum of near-independent SNPs obtained from approximate COJO multi-SNP analyses of a fixed-effect meta-analysis of all contributing GWASs. The multi-ancestral PGS_LC reflects the best-performing PRS-CSx score consisting of a linear combination of five ancestry-specific scores, with weights being specific to the validation population (for example, AFR). Population labels follow PAN-UKBB assignment of genetically determined ancestry. Sample sizes (distinct individuals): African 6,154; Admixed American 971; Middle Eastern 1,553; East Asian 2,660; Central/South Asian 8,005; European 20,000. b, Explained variance for BMI within validation populations, comparing the multi-ancestry PGS_LC to a previously published score (PGS_Khera) based on a smaller BMI GWAS meta-analysis. Same R² definition and CI estimation as in a. Population descriptors reflect a combination of self-identified ethnicity and genetic similarity. For the MVP’s non-Hispanic Asian (AS) group, the result shown is for the PGS_LC using the linear combination weights derived from UKBB-EAS. Sample sizes (distinct individuals), from left to right: AFR 12,263, 2,332, 18,701; AMR 10,281, 8,096; AS 4,201; EAS 1,359; SAS 1,177; EUR 13,673, 69,828, 340,224. c, Separation in BMI, body fat percentage (BF%) and waist-to-hip ratio (WHR) across deciles of the PGS_LC within the validation subset of the UKBB participants of European-like ancestry (n ~ 340,000). All traits were rank-based inverse-normal transformed by sex.

Fig. 3. Prediction of prevalent obesity outcomes in adults.

a, Separation in prevalence of obesity (BMI ≥ 30 kg m⁻²) across 1% groups of PGS_Khera and PGS_LC within the validation subset of the UKBB participants of European-like ancestry (n ~ 340,000), with reference lines for the bottom and top 1% groups. Error bars show 95% confidence intervals (CIs) based on the normal approximation to the binomial distribution. The horizontal lines correspond to the average prevalence (black, dotted) and the prevalence of obesity within the top and bottom 1% of PGS_Khera and PGS_LC (red and blue, respectively). b, Odds ratios with 95% CIs for prevalent obesity class I or higher, per s.d. of PGS_Khera and PGS_LC, adjusted for age, sex, principal components of ancestry and genotyping array. All PGSs were standardized using the mean and s.d. of the PGS within individuals who did not have obesity class I or higher, to account for differences in prevalence across validation populations. Sample sizes (distinct individuals), from left to right: AFR 12,263, 2,332, 18,701; AMR 10,281, 8,096; AS 4,201; EAS 1,359; SAS 1,177; EUR 13,673, 69,828, 340,224. c, AUC classification of prevalent obesity outcomes in the BioMe Biobank, the MVP and the UKBB. Models including PGSs (included as a continuous predictor) additionally include principal components of ancestry. CBS10, self-reported comparative body size at age 10 years. Restricted to estimates where the number of individuals with the obesity outcome was at least 50.

Fig. 4. PGS_LC performance during childhood and adolescence.

a, Repeated cross-sectional linear regression associations of standardized PGS_LC with BMI and height, with both standardized within sample by sex and timepoint. Data are presented as regression coefficient with 95% confidence interval (CI). Ponderal index was used instead of BMI at birth. Associations were adjusted for age and principal components of ancestry. Sample sizes, based on repeated measurements, from left to right, for BMI: 4,740, 638, 847, 814, 769, 41, 732, 729, 725, 720, 699, 5,816, 4,863, 5,570, 5,368, 5,187, 4,910, 4,556, 4,024, 3,603, 2,780; and height: 4,802, 638, 847, 814, 771, 741, 734, 729, 726, 723, 701, 5,820, 5,160, 5,572, 5,379, 5,188, 4,956, 4,561, 4,032, 3,606, 2,782. b, Sequentially plotted mean BMI trajectories from the age of 4 months to 24 years with knot points from linear spline multilevel models, accounting for sex and principal components of ancestry, of PGS_LC (bottom 10%, middle 80%, top 10%). c, Contribution of PGS_LC to explained variance (adjusted R²) for BMI, rank-based inverse-normal transformed by sex and timepoint. Ponderal index was used instead of BMI at birth. Data are presented as R² values computed from the original dataset, with error bars representing 95% confidence intervals (2.5th–97.5th percentiles) estimated from 1,000 bootstrap resamples. Predictors available at birth were birthweight, maternal education, pre-pregnancy maternal BMI, maternal age at date of birth and household social status. The left panel shows explained variance for BMI at the timepoint shown on the x axis. In contrast, the right panel shows explained variance for BMI measured at 18 years, with early-life BMI measurement shown on the x axis used as predictors. Sample sizes, based on repeated measurements, from left to right: 3,800, 839, 714, 1,192, 4,062, 3,044, 940, 594, 737, 725, 890, 3,310.

Fig. 5. AUC for obesity classification at age 50, stratified by having overweight or obesity at age 20.

Analyses performed within the PLCO Cancer Screening Trial. Restricted to estimates where the number of individuals with the obesity outcome was at least 50. Population descriptors were provided by the PLCO investigators and reflect genetically determined ancestry using Genetic Relationship and Fingerprinting (GRAF). PCs, principal components; y, years; M1, model with birth year, sex, BMI at 20y, and PCs.

Fig. 6. Impact of PGS_LC on weight change due to ILIs.

Interaction effect between PGS_LC and trial arm (ILI versus comparison arm) for weight change within the first year of follow-up (left) and weight change after the first year in the subset of individuals who had lost ≥3% of baseline weight at year 1 (right). Data are presented as study-specific and study-combined interaction regression coefficients (with 95% confidence interval (CI)), pooled through inverse-variance weighted fixed-effect meta-analysis. Weight change was assessed by using weight at follow-up timepoint(s) as outcome in the linear mixed models, adjusting for initial weight (baseline or year 1, respectively). LA, Look AHEAD. Population descriptors reflect race and ethnicity, as reported by dbGAP variable phv00201855.v2.p1 (DPP) or self-report by participants (LA). Study-specific sample sizes (distinct individuals, with pooled sample sizes for ‘All’), from top to bottom, for forest plot on the left: 295, 374, 251, 428, 839, 1,722, 1,385, 2,524; and on the right: 118, 186, 128, 233, 428, 951, 674, 1,370.

Extended Data Fig. 1. Distribution of obesity categories and average BMI in adult tuning and validation populations.

Distribution of obesity categories and mean BMI (+/- standard deviation) across adult tuning and validation populations. Asian-specific BMI cutoffs (see Methods) were applied to East Asian (EAS), South Asian (SAS), and non-Hispanic Asian (AS) populations, in contrast to African (AFR), American (AMR), European (EUR), and Middle Eastern (MID) populations. UKBB, UK Biobank; MVP, the Million Veteran Program; GPC-UGR, the Uganda General Population Cohort. Sample sizes (distinct individuals), from left to right, for Tuning: 6154, 971, 1553, 2660, 8005, 20000; Validation: 12263, 2332, 18701, 10281, 8096, 4201, 1359, 1177, 13673, 69828, 340224; and BMI: 12263, 10281, 1359, 1177, 13673, 2332, 18701, 8096, 4201, 69828, 6154, 971, 1553, 2660, 8005, 340224.

Extended Data Fig. 2. Prevalence of obesity categories across 1%-groups of PGSs.

Prevalence of obesity category, with 95% confidence intervals based on the normal approximation to the binomial distribution, across 1%-groups of PGS_Khera and PGS_LC within the validation subset of the UKBB participants of European ancestry (N ~ 340k). The black dotted line denotes the average prevalence in the entire subset. The red and blue lines correspond to the prevalence of the obesity category within the top and bottom 1% of PGS_Khera and PGS_LC, respectively.

Extended Data Fig. 3. Odds ratio per standard deviation (SD) increase of PGS_LC for prevalent obesity categories.

All PGS were standardised using the mean and SD of the PGS of individuals who did not have the outcome of interest (that is, those with a BMI value below the threshold for a given obesity category), to account for differences in prevalence across validation populations. MVP, Million Veteran Program; GPC-UGR, Uganda General Population Cohort; UKB, UK Biobank. Sample sizes (distinct individuals) per grouped set of estimates, left to right: AFR 12263, 2332, 18701; AMR 10281, 8096; AS 4201; EAS 1359; SAS 1177; EUR 13673, 69828, 340224.

Extended Data Fig. 4. Distributional shifts of ancestry-specific scores before and after ancestry-calibration.

Distributional shifts of ancestry-specific scores which are linearly combined to create PGS_LC, across UK Biobank-tuning and BioMe-validation populations, before (top) and after (bottom) ancestry–calibration using 1000 Genomes reference data (Methods). Each plot shows a single ancestry-specific score, with separate lines showing the distribution across ancestries. For these plots, each score was standardised to mean zero and variance one once across all participants of a given study.

Extended Data Fig. 5. Association between PGS_LC and height trajectory between birth and age 24.

Estimated using linear spline multilevel models with repeated measures from the ALSPAC cohort.

Extended Data Fig. 6. AUC for obesity outcomes at age 50, stratified by sex and overweight/obesity status at age 20.

Estimated in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial. Restricted to estimates where the number of individuals with the obesity outcome was at least 50. Population labels as provided by PLCO.

Extended Data Fig. 7. Distribution of PGS across intensive lifestyle intervention trial arms.

Scores were standardised to mean zero and variance one by population group, ignoring assigned trial arm. Vertical lines indicate the mean PGS per arm.